Statement on US government directive to suspend access to Fable 5 and Mythos 5
Points and comments are a snapshot, not live.
US government orders Anthropic to suspend all access to Fable 5 and Mythos 5 models citing national security concerns over a jailbreak vulnerability.
Anthropic received a government directive on January 28 at 5:21pm ET requiring suspension of Fable 5 and Mythos 5 for all users, including foreign nationals and non-US employees. The government cited discovery of a method to bypass the model's safeguards. Anthropic reviewed the demonstrated jailbreak, which consists of asking the model to read a codebase and fix software flaws, and concluded the capability is widely available in other models including OpenAI's GPT-5.5 and is used daily by cybersecurity defenders. Anthropic states it conducted thousands of hours of red-teaming with government agencies and third parties before launch, found no universal jailbreak, and adopted a defense-in-depth strategy acknowledging that perfect jailbreak resistance is not currently achievable. The company is complying with the legal directive but disagrees that a narrow jailbreak warrants recalling a commercially deployed model, arguing the standard would halt all frontier model deployments across the industry.
What commenters are saying
Commenters split between viewing this as effective marketing for Anthropic versus legitimate regulatory overreach. Some note the jailbreak capability exists in competing models like GPT-5.5 and questioned why only Anthropic faced the directive. Others accused Anthropic of overhyping model dangers to drive regulatory capture and market dominance. Practical concerns surfaced around refunds for affected subscribers. A dominant critique argues Anthropic's public fear-mongering about AI risks, combined with this incident, suggests the company engineered regulatory advantage rather than facing genuine security discovery.